iguanid lizard (Oplurus quadrimaculatus)
Ole Theisinger, Wiebke Berg & Kathrin H. Dausmann
Dept. of Functional Ecology, Zoological Institute, University of Hamburg, 20146 Hamburg, Germany.
In press at the Journal of Thermal Biology, DOI: 10.1016/j.jtherbio.2017.01.005 Abstract
Physiological or behavioural adjustments are a prerequisite for ectotherms to cope with different thermal environments. One of the world’s steepest environmental gradients in temperature and precipitation can be found in southeastern Madagascar. This unique gradient allowed us to study the compensation of thermal constraints in the heliothermic lizard Oplurus quadrimaculatus on a very small geographic scale. The lizard occurs from hot spiny forest to intermediate gallery and transitional forest to cooler rain forest and we investigated whether these habitat differences are compensated behaviourally or physiologically. To study activity skin temperature (as proxy for body temperature) and the activity time of lizards, we attached temperature loggers to individuals in three different habitats. In addition, we calculated field resting costs from field resting metabolic rate to compare energy expenditure along the environmental gradient. We found no variation in activity skin temperature, despite significant differences in operative environmental temperature among habitats. However, daily activity time and field resting costs were reduced by 35 % and 28 % in the cool rain forest compared to the hot spiny forest. Our study shows that O. quadrimaculatus relies on behavioural mechanisms rather than physiological adjustments to compensate thermal differences between habitats.
Furthermore, its foraging activity in open, sun exposed habitats facilitates such a highly effective thermoregulation that cold operative temperature, not energetically expensive heat, presents a greater challenge for these lizards despite living in a hot environment.
28
Thermal constraints along an environmental gradient Keywords
Behavioural compensation; energy expenditure; metabolic rate; operative environmental temperature; physiological adjustment; reptiles.
Introduction
Species’ distribution limits are defined by various parameters (Guisan and Thuiller 2005).
In ectotherms, the distribution of operative environmental temperature (Te), i.e. the range of attainable body temperatures (Tb), is one of the key factors (Angilletta 2009; Huey et al. 2012; Sinervo et al. 2010). Even efficient and precise thermoregulation, as found in heliothermic lizards, still depends on Te to provide suitable Tb (Bakken and Angilletta 2014). If Te is too high, lizards must retreat to cool refuges to avoid overheating. If Te is too low, the animals’ performance is reduced. Sinervo et al. (2010), for example, predicted that Mexican lizards are prone to extinction because of increasing ambient temperature and prolonged hours of restriction (i.e. periods of forced inactivity due to unsuitably high ambient temperature). Tropical lizards are thought to show narrow thermal tolerances because they have adapted to relatively stable climates (Ghalambor et al. 2006; Huey et al. 2009; Janzen 1967). However, recent results contradict this assumption (Leal and Gunderson 2012). A variety of microclimates, altitudinal gradients, and pronounced seasonality provide a wide range of environmental conditions in the tropics (Dewar and Richard 2007) and thus might have led to higher thermal tolerances than assumed so far.
The prerequisite for ectotherms to occur in differing thermal environments are physiological or behavioural adjustments. Physiological adjustments comprise changes in physiological rates (Seebacher et al. 2015) or shifts in the thermal tolerance of a species (Gunderson and Stillman 2015). Furthermore, the preferred body temperature [Tpref; the selected Tb of undisturbed ectotherms in an artificial temperature gradient (Huey and Kingsolver 1989)] and the physiological optimum can be lowered or increased in accordance with ambient conditions (Blouin-Demers et al. 2000; Clusella-Trullas and Chown 2014). An upshift of the Tb set-point as a response to higher ambient temperature, for example, increases the potential activity time and might reduce costs for thermoregulation in the face of climate change (Gvozdik 2012). Despite Tb selection
29 Thermal constraints along an environmental gradient
being a behavioural response, the driver for this adjustment are changes in the thermal reaction norm and hence of physiological nature (Little and Seebacher 2016). Lizards that are exposed to different temperatures over several weeks in the laboratory show acclimation of Tpref (Blumberg et al. 2002). This effect was also observed under natural conditions between seasons (Diaz et al. 2005; Seebacher and Grigg 1997) but there is no or only little evidence for adjustments of Tpref to different habitats and altitudes (Gvozdik and Castilla 2001; Van Damme et al. 1989). However, this apparent absence of physiological adjustments is not necessarily equivalent to an incapability of lizards to acclimatize.
Other mechanisms, such as changes in thermoregulatory behaviour and adjusted daily activity time, are known to be effective (Gvozdik 2002) and might be preferential to physiological acclimatization under natural circumstances (Gunderson and Stillman 2015). Thermal constraints through high ambient temperature and the effect of climate warming have often been studied (Deutsch et al. 2008; Huey et al. 2009; Kearney et al.
2009; Sinervo et al. 2010). Since metabolic rate increases exponentially with Tb, climate warming might lead to excessive energetic costs (Kearney 2013). These can be assessed by calculating maintenance costs, i.e. basic metabolic costs in fasting and resting ectotherms, using field Tb (or a proxy for field Tb) and the corresponding metabolic rate (Kearney and Porter 2004). However, low ambient temperature can also be challenging, especially for warm-adapted species (Grbac and Bauwens 2001; Van Damme et al. 1987).
Lower Tb may be beneficial for reducing energy expenditure (Christian et al. 1996; Huey et al. 1989) but sprint speed, digestion, nutrient assimilation and other physiological functions may be negatively affected by suboptimal performance (Angilletta et al. 2002;
Huey and Kingsolver 1989).
To study these compensatory mechanisms under natural conditions, we chose one of the world’s steepest environmental gradients, which is found in southeastern Madagascar.
The unique gradient in ambient temperature, humidity and rainfall, connects dry spiny forest and humid rain forest via transitional forest and gallery forest along rivers within less than 5 km (Goodman 1999). Due to this very small geographic scale, species could potentially move between all habitats within a lifetime or sooner, depending on mobility, and reptile species from the hot and dry spiny forest can also be found in the cooler rain forest (pers. obs.).
30
Thermal constraints along an environmental gradient
Duméril’s Madagascar swift (Oplurus quadrimaculatus) is one of these species. It occurs in a wide range of habitats with strikingly different environmental conditions including the climatic extremes of hot, dry spiny forest and humid, cool rain forest. We analyzed daily skin temperature (Tskin) patterns (as a proxy for Tb; Berg et al. 2015) in individuals across the entire hot to cold environmental gradient including all four habitat types. Our aim was to examine if these lizards compensate thermal differences through adjustments in activity Tskin or through differences in daily activity time. In addition, we measured field resting metabolic rate (field RMR) and compared individuals’ field resting costs (maintenance costs of non-fasting lizards at daytime) to assess the effect of differing thermal environments on basic energy needs.
Methods
Study sites
Our study sites were located in the Andohahela National Park (24°57'S, 46°35'E) on the western slope of the Anosy Mountains in southeast Madagascar. Within our study area, the distance between the two most divergent habitats (spiny forest and rain forest) is less than 5 km. The spiny forest (150 – 160 m a.s.l.) is characterized by scant and xerophile vegetation such as the octopus tree (Dideracea spp.) and the evergreen rain forest (400 – 430 m a.s.l.) consists of large shady trees and dense understory. The transitional forest (280 – 380 m a.s.l.), as the name implies, comprises mixed vegetation with similar moderate environmental conditions as the gallery forest along rivers (130 – 140 m a.s.l.;
Andriaharimalala et al. 2011).
Study species
Oplurus quadrimaculatus is a heliothermic iguana with a body mass of (mean ± SD) 76.5
± 10.5 g (N = 310) and a snout-vent-length of 12.8 ± 0.6 cm. It inhabits open rocky habitats and is a sit-and-wait forager that feeds mainly on flying insects. Its main distribution is the spiny forest of southern Madagascar but there are also populations at higher altitudes and more humid habitats (Glaw and Vences 2007). This species is highly philopatric with a small home range of sometimes less than 50 m2. Hence, animals can easily be located for recapture.
31 Thermal constraints along an environmental gradient
Operative environmental temperature
We used specifically designed copper models (three models per habitat) that matched the lizards in shape and colour to characterize the thermal habitat of individual O.
quadrimaculatus (Dzialowski 2005). We equipped these models with temperature loggers (Thermochron iButtons, model DS1921G; resolution ± 0.5°C; weight 3.3 g; Maxim Integrated Products Inc., San Jose, California, USA.; calibrated in a water bath before use). Calibration against live animals showed that filling the copper models with fine sand revealed the best correlation (Pearson correlation coefficient = 0.962; p < 0.001) with no differences in heating and cooling rates between copper models and lizards (t33 = 0.857; p
= 0.397). After programming the loggers to record the core temperature of the model every five minutes, we distributed three models in each habitat (full sun, shade and crevice) in order to cover a representative Te-range. Measurements were made for six consecutive days in each habitat on cloudless, sunny days to ensure similar abiotic conditions.
Skin temperature patterns
To measure daily Tskin patterns, we noosed 48 individuals from different sites across the designated habitats during the rainy season (Mid-October to Mid-March) of 2009/2010, 2010/2011 and 2011/2012. We assume a similar sex-ratio in each habitat because this species is highly philopatric with a very small home range and it occurs in pairs.
However, the certain identification of sex was not always possible and thus sex is not taken into account. We glued calibrated temperature loggers on the animals’ backs and released them at their points of capture (Fig. S1.1). To make sure that the weight of the device did not exceed the recommended 5 % of the animals’ body weight (Lovegrove 2009), we only equipped adult individuals with a minimum body mass of 70 g with a temperature logger. We used superglue (UHU Sekundenkleber, UHU GmbH, Bühl, Germany) to attach the devices. The recording interval was set to five minutes and the temperature loggers were able to store up to 2084 data points, which led to a maximal measuring time of seven continuous days. We recaptured individuals after five to seven days to remove the loggers. We gained Tskin data from 13 individuals in the dry spiny forest, 25 individuals in the moderate gallery and transitional forest, and 10 individuals in the humid rain fores. Tskin is highly related to Tb and it can be used as a substitute in this species (Berg et al. 2015). In contrast to day Tskin, which includes the entire photoperiod
32
Thermal constraints along an environmental gradient
from 0600 hours until 1800 hours, activity Tskin is defined as the time-span from the first Tskin peak in the early morning after initial heating until a drop in Tskin indicating retreat to a crevice, which is clearly visible in distinct daily Tskin patterns (Fig. 1.1). As cloud cover and rain may lead to behavioural changes (Sun et al. 2001), we only used sunny, cloudless days in our analysis. We also installed temperature loggers on trees in 1 m height in full shade and protected from wind and rain in the same habitat to measure ambient temperature. These measurements served as reference for sudden weather changes.
Metabolic measurements
To measure the oxygen consumption and to calculate metabolic rate of O.
quadrimaculatus, we used a portable open flow oxygen analyzer (OxBox, designed and constructed by T. Ruf and T. Paumann, FIWI, University of Veterinary Medicine Vienna, Austria) with fuel cell oxygen sensors (7OX-V CiTiceL; City Technology Ltd, Portsmouth, UK; accuracy < 0.02 vol. %). The device was calibrated using a gas mixing pump (type G27, Wölsthoff, Bochum, Germany) each time before and after the field season. We measured non-fasting, resting but awake animals and therefore refer to our measurements as field RMR and use these data to estimate energy budgets close to natural conditions (Niewiarowski & Waldschmidt, 1992; Levesque et al. 2016). We captured adult lizards (n = 35) in the morning between 0900 hours and 1200 hours, weighed and placed them in a plastic container of 1 l volume that served as metabolic chamber. We used a pull-mode respirometry setup and recorded sample air once per minute at an airflow of 40 l/h alternated with hourly six-minute samples from ambient reference air for baseline corrections. Sample and reference air was dried and cleaned through silica-scrubber before the measurement. Measurements lasted from early morning until 1600 hours. The temperature in the metabolic chamber was monitored with thermocouples to ensure smooth heating curves. Individuals showing signs of heat stress were immediately removed from the metabolic chamber. In addition, the lizards’ Tskin was recorded via attached temperature loggers and ranged from 17.5°C to 42.5°C. We calculated VO2 and corrected for CO2 as recommended by Lighton (2008). Oxygen consumption (ml O2/(g*h)) was then calculated with the following equation: field RMR = flow * ∆ vol % O2 * 10. Finally, oxygen consumption (ml O2/(g*h)) was converted into energy expenditure (1 ml O2/(g*h) = 20.08 J/(g*h)). Field RMR values were then
33 Thermal constraints along an environmental gradient
assigned to the Tskin of the animal. We used 5-min-averages and manually excluded periods of animal activity. We obtained between 7 and 75 field RMR values per individual (25.9 ± 17.5 records/individual) for a total of 35 individuals distributed equally across habitats. After the measurement, the animals were weighed again and released at the original place of capture.
Daytime field resting costs
For the calculation of field resting costs (J/(g*12h)) we first calculated the temperature-rate relationship of Tskin and field RMR using a linear mixed effects model (LMM). These models account for an unequal number of data points per individual and dependent data from the same individual. We used the lme function in the R package nlme (Pinheiro et al. 2015) and included Tb as covariate and individual ID as a random factor. We visually checked the model residuals with Q-Q plots and histograms to assess the model quality.
All analyses were performed in R v3.2.1 (R Development Core Team 2015). Using the resulting equation, we then calculated corresponding field RMR values for each Tskin measurement of the daily Tskin pattern (measured in 5 min intervals) and summed these across the photoperiod to obtain field resting costs (J/(g*12h)). Statistical analyses were conducted with IBM SPSS 21.0. We calculated an ANOVA to test for differences in activity Tskin, activity time and field resting costs. The number of measured days per individual varied between one and five days. We therefore used mean values per individual for a comparison among habitats. All means are given with standard deviation.
Results
Mean activity Tskin was 37.5 ± 0.8 °C (n = 48) and did not differ among habitats (Fig. 1.2;
ANOVA: F2;45 = 0.168; p = 0.846) but the mean day Tskin during the photoperiod, which also includes hours of restriction due to constraints in Te, decreased from spiny forest to rain forest (ANOVA: F2;45 = 59.725; p < 0.001). Tskin during the night (between 1800 hours and 0600 hours), when the lizards are inactive, also decreased from spiny to rain forest (ANOVA: F2;45 = 47.427; p < 0.001). Minimum and maximum Te as well as Te -range decreased along the gradient and the duration of temperatures in the activity Tskin
range was shorter in the rain forest (Fig. 1.3). The fluctuating Te measured in the sun in the rain forest was caused by shading from branches but sunny spots were constantly
34
Thermal constraints along an environmental gradient
available throughout the day and hence, Te was constantly high between these peaks.
Differences in daily activity time despite similar activity Tskin are shown using exemplary Tskin patterns from individuals across the environmental gradient (Fig. 1.1).
Figure 1.1: Daily skin temperature (Tskin) profiles of Oplurus quadrimaculatus across an environmental gradient in southeast Madagascar. The solid line shows Tskin of the lizard and the dashed line shows ambient temperature. The lizard 1) leaves its warm crevice and cools down to ambient temperature before heating up in the sun by basking, 2) is active, 3) cools down with ambient temperature and enters its crevice.
Horizontal black bars indicate the scotophase.
Figure 1.2: Mean activity skin temperature (Tskin; open circles), mean day Tskin during the photopase (between 0600 hours and 1800 hours) including periods of inactivity (grey squares), and night Tskin (black triangles) of Oplurus quadrimaculatus in different habitats along an environmental gradient in southeast Madagascar. Error-bars show 95% confidence intervals and lowercase letters (a, b, c and x, y, z) indicate significant differences between habitats.
35 Thermal constraints along an environmental gradient
Figure 1.3: Mean operative environmental temperature of Oplurus quadrimaculatus in different habitats along an environmental gradient in southeast Madagascar measured with copper models during the same time period. The straight black line shows mean activity skin temperature in each habitat and straight grey lines are standard deviation. Horizontal black bars indicate the scotophase.
The daily activity time differed among habitats (ANOVA: F2;45 = 8.582; p < 0.001). We found a decline in activity time from the hot spiny forest through to the cooler rain forest (Fig. 1.4A). While lizards in the spiny forest were active for 464 ± 60 min per day (n = 13), individuals in the moderate gallery and transitional forest were active for 375 ± 103min (n = 25), and individuals in the rain forest for 300 ± 108 min (n = 10). The duration of activity in the rain forest thus corresponds to only 65 % of the time spent active in the spiny forest.
The temperature-field RMR relationship is described by log(field RMR) = 0.09(Tskin) – 5.37 (Fig. 1.5). We found differences in the daytime field resting costs of individuals among habitats (Fig. 1.4B; ANOVA: F2;45 = 20.885; p < 0.001). The highest field resting costs were found in the spiny forest (68.9 ± 6.4 J/(g*12h); n = 13) followed by the gallery and transitional forest (61.1 ± 5.5 J(g*12h); n = 25) and the rain forest (49.8 ± 10.5 J/(g*12h); n = 10). This implies a reduction in field resting costs of 28 % from the spiny forest to the rain forest.
36
Thermal constraints along an environmental gradient
Figure 1.4: A) Mean daily activity time (DAT) and B) daytime field resting costs (FRC) of Oplurus quadrimaculatus in different habitats along an environmental gradient in southeast Madagascar. Error-bars show 95% confidence intervals and lowercase letters indicate statistical significance.
Discussion
Our data demonstrate that O. quadrimaculatus is a precise thermoregulator that maintains a high activity Tskin. While activity time differs significantly between habitats, this species shows no adjustments in activity Tskin between different thermal environments. During activity, Tb can be affected by several biotic and abiotic factors, such as food quantity, habitat structure and predation (Clusella-Trullas and Chown 2014). However, O.
quadrimaculatus thermoregulates with high precision despite marked thermal differences between habitats and maintains a narrow range of activity Tskin throughout the day despite
37 Thermal constraints along an environmental gradient
a much wider range of Te. It therefore seems reasonable to assume that the observed activity Tskin is closely related to the temperature set-point of these lizards and hence, that the thermal optimum does not vary along the environmental gradient. In accordance with other field studies, which failed to find adjustments of activity Tb or Tpref between different thermal environments (Brown and Griffin 2005; Gvozdik and Castilla 2001;
Van Damme et al. 1989), this could imply that differences in the thermal environment are not sufficient for provoking physiological acclimatization because they are counterbalanced through behavioural thermoregulation. Plasticity of Tpref, either under laboratory conditions or seasonally in the natural environment, seems to occur only if the animals’ Tb is constantly forced away from Tpref (Blumberg et al. 2002; Diaz et al. 2005;
Seebacher and Grigg 1997), presumably because physiological acclimatization is also accompanied by energetic costs through increased rates of transcription of regulatory enzymes (Seebacher 2005). The precise thermoregulation to a narrow range of activity Tskin across habitats and the absence of metabolic responses along the gradient (Berg et al.
in prep.) indicate that behavioural mechanisms are sufficient to compensate for thermal differences.
Figure 1.5: Temperature-field resting metabolic rate (field RMR) relationship for Oplurus quadrimaculatus.
Data points represent repeated measurements from all individuals and the black line shows the fitted line of the linear mixed effects model (log(field RMR) = 0.09(Tskin) – 5.37) that accounts for an unequal number of data points and repeated measurements.
Nevertheless, metabolic rate often shows acclimatization in species that are active year-round (Seebacher 2005). Seasonal metabolic acclimatization suggests that O.
quadrimaculatus exhibits physiological compensation rather as a supplementary
38
Thermal constraints along an environmental gradient
mechanism if behavioural thermoregulation is limited. This acclimatization response is restricted to lower Tskin because increased rates at activity Tskin would result in overcompensation, which means that energy expenditure would be unnecessarily high.
These data reveal that this species exhibits physiological acclimatization to compensate seasonal thermal challenges (Berg et al. in prep.) but physiological adjustments to the different habitats along the gradient are not necessary.
The daily activity time of O. quadrimaculatus declines from hot spiny forest to cooler rain forest by 35 %. Lizards in the rain forest can still reach their activity Tskin through basking but the window of suitable Te is shorter and the overall activity time (between the first rise in Te and the final decrease in Te) is therefore reduced. Surprisingly, we did not find heat-induced hours of restriction in the hot spiny forest as has been described for numerous Mexican lizard species in similarly hot habitats (Sinervo et al. 2010).
Considering that we only used Tskin patterns from cloudless, sunny days, one would expect that our measured Te is higher than average but even in the hot spiny forest, the lizards did not seem to reach their upper thermal limits. This may be due to the heterogeneity of the habitat, which can be as important as mean ambient temperature for the animal’s performance (Sears and Angilletta 2015). Despite being open rocky plateaus, the habitat provides sufficient spacial structure and shady spots to avoid overheating at increased but apparently not excessive costs for behavioural thermoregulation. Moreover, a constantly high Tskin seems to be important for O. quadrimaculatus: either lizards maintain a minimum Tskin of 36 °C or they retreat to crevices. No trade-off in the form of activity at lower Tskin has been observed, at least on sunny days. We assume that performance drops significantly at Tskin below 36 °C, making it difficult to perform quick targeted bursts to catch the preferred flying prey and reducing their ability to escape from predators (Cooper 2000). This indicates that the challenging factor for O.
quadrimaculatus is not to keep cool but to stay hot, which is more difficult in the cool, shady rain forest. Hence, we only found thermal constraints at the cold end of this species’ distribution range.
In our study, the field resting costs in the rain forest were 28 % lower than those in the hot spiny forest. Thus, field resting costs decline in a similar way to daily activity time, which was reduced by 35 % in the cooler habitat. We suggest two main reasons for this pattern.
First, less time at high activity Tskin reduces the overall metabolic costs. Second, Te, and
39 Thermal constraints along an environmental gradient
crevice temperature in particular, was lower in the rain forest, even during the photoperiod. Hence, when the lizards were thermoconform during hours of restriction (induced by low Te), the field RMR was passively reduced. Our calculations are based on the effect of Tskin on mean field RMR and do not include costs of locomotion or possible seasonal acclimatization effects. One can argue that metabolic rate can be threefold higher during activity (Bennett and Nagy 1977). However O. quadrimaculatus is a sit-and-wait forager that performs only few targeted bursts for prey, so costs for physical activity adding to field RMR are relatively low. Field resting costs are thus likely to be close to the actual energy expenditure during the day. Since we did not measure standard metabolic rate during the night, we only calculated field resting costs during the photoperiod. However, the high mean night Tskin (31.7 ± 1.8 °C) in the spiny forest might be especially costly because at high temperatures even small increases in Tskin lead to a substantial increase in energy expenditure. Christian et al. (1996), for example, found that a decrease of 4 °C in night-time Tb resulted in a decrease of energetic costs of more than 30 % for the water monitor Varanus mertensi. Hence, lower night temperature in the rain forest might amplify the energy savings of O. quadrimaculatus.
Why is the cooler rain forest habitat more challenging if the lizards can save so much energy? Most striking is that the time suitable for activity is highly reduced, which includes feeding, mating, and territorial behaviour. Consequently, these lizards must feed and digest in a significantly shorter time period. The most challenging restriction is probably that low Tb also depresses biochemical processes that support digestion and energy assimilation. Even if the lizards find enough food within the limited time frame, digestion might suffer from low Tb as shown for the eastern fence lizard Sceloporus undulatus (Angilletta et al. 2002). This effect will be even more pronounced if one considers the higher number of cloudy and rainy days and cooler wind in the rain forest, where maximum Te is below the threshold for activity and suboptimal for other physiological functions. In fact, while our study site in the rain forest is the eastern distributional limit of O. quadrimaculatus, the species occurs in even drier and hotter areas towards the southwest of Madagascar (Glaw and Vences 2007). Adult lizards have a wide spectrum of mechanisms to cope with different environmental temperatures but egg development and juvenile growth, which are especially sensitive to temperature, are largely understudied (Du and Shine 2015). Apart from potential movements within the egg (Du et al. 2011), the embryos’ Tb depends largely on the environment. Cold
40
Thermal constraints along an environmental gradient
temperature lead to prolonged incubation periods, embryonic diapause and maybe reduced hatchling survival rate (Du and Shine 2015). Considering the distribution in the hot spiny forest and the high activity Tb of adult O. quadrimaculatus, the rain forest possibly acts as a population sink.
In summary, our study shows that heliothermic lizards strongly depend on behavioural mechanisms for coping with varying environmental conditions. Highly effective thermoregulatory behaviour and adjustments in activity time are the prevalent mechanisms and thus, low temperatures rather than hot environments seem to be challenging for these species. Cooler habitats may reduce the energy expenditure in warm adapted species but the consequences of decreased physiological functioning and reduced time for food intake remain unclear. Hence, heliothermic lizards may even profit - at least initially - from an enlarged distribution range caused by climate warming and by increased basking opportunities through forest degradation. We found that behavioural thermoregulation is sufficient to compensate extreme climatic variation and the potential of this mechanism may render open habitat species less vulnerable to climate change compared to species with limited capacity for behavioural adjustments.
Acknowledgements
The project was carried out under the Accord de Collaboration between the University of Antananarivo and the University of Hamburg. We thank Jörg Ganzhorn for helpful advice in the preparation of this study. We are grateful to James Turner and to the anonymous reviewers for helpful comments on our manuscript and to Jacques Rakotondranary for assistance in organizing paper work and field logistics in Madagascar. We also thank local field assistants for their patience and endurance during long field work days. This study was funded by the German Academic Exchange Service (DAAD) and Evangelisches Studienwerk Villigst e.V. Research was conducted under Permit #113/09,
#205/11 and #046/12 from Madagascar National Parks and the Ministère des Eaux et Forêts Madagascar.
41 Thermal constraints along an environmental gradient
References
Andriaharimalala, T., Roger, E., Rajeriarison, C., Ganzhorn, J., 2011. Analyse structurale des différents types de formation végétale du Parc National d’Andohahela (Madagascar) comme habitat des animaux. Malagasy Nat 5, 39–58.
Angilletta, M.J., 2009. Thermal adaptation: a theoretical and empirical synthesis. Oxford University Press, Oxford, UK.
Angilletta, M.J., Hill, T., Robson, M.A., 2002. Is physiological performance optimized by thermoregulatory behavior?: a case study of the eastern fence lizard, Sceloporus undulatus. J Therm Biol 27, 199–204.
Bakken, G.S., Angilletta, M.J., 2014. How to avoid errors when quantifying thermal environments. Funct Ecol 28, 96–107.
Bennett, A.F., Nagy, K.A., 1977. Energy-expenditure in free-ranging lizards. Ecology 58, 697–700.
Berg, W., Theisinger, O., Dausmann, K.H., 2015. Evaluation of skin temperature measurements as suitable surrogates of body temperature in lizards under field conditions. Herpetol Rev 46, 157–161.
Berg, W., Theisinger, O., Dausmann, K.H. (in prep.) Acclimatization patterns in tropical reptiles: uncoupling temperature and energetics.
Blouin-Demers, G., Kissner, K.J., Weatherhead, P.J., 2000. Plasticity in preferred body temperature of young snakes in response to temperature during development.
Copeia, 841–845.
Blumberg, M.S., Lewis, S.J., Sokoloff, G., 2002. Incubation temperature modulates post-hatching thermoregulatory behavior in the Madagascar ground gecko, Paroedura pictus. J Exp Biol 205, 2777–2784.
Brown, R.P., Griffin, S., 2005. Lower selected body temperatures after food deprivation in the lizard Anolis carolinensis. J Therm Biol 30, 79–83.
42
Thermal constraints along an environmental gradient
Christian, K.A., Weavers, B.W., Green, B., Bedford, G.S., 1996. Energetics and water flux in a semiaquatic lizard, Varanus mertensi. Copeia, 354–362.
Clusella-Trullas, S., Chown, S.L., 2014. Lizard thermal trait variation at multiple scales: a review. J Comp Physiol B 184, 5–21.
Cooper, W.E., 2000. Effect of temperature on escape behaviour by an ectothermic vertebrate, the keeled earless lizard (Holbrookia propinqua). Behaviour 137, 1299–1315.
Deutsch, C.A., Tewksbury, J.J., Huey, R.B., Sheldon, K.S., Ghalambor, C.K., Haak, D.C., Martin, P.R., 2008. Impacts of climate warming on terrestrial ectotherms across latitude. P Natl Acad Sci USA 105, 6668–6672.
Dewar, R.E., Richard, A.F., 2007. Evolution in the hypervariable environment of Madagascar. P Natl Acad Sci USA 104, 13723–13727.
Diaz, J.A., Cabezas-Diaz, S., Salvador, A., 2005. Seasonal changes in the thermal environment do not affect microhabitat selection by Psammodromus algirus lizards. Herpetol J 15, 295–298.
Du, W.G., Shine, R., 2015. The behavioural and physiological strategies of bird and reptile embryos in response to unpredictable variation in nest temperature. Biol Rev 90, 19–30.
Du, W.G., Zhao, B., Chen, Y., Shine, R., 2011. Behavioral thermoregulation by turtle embryos. P Natl Acad Sci USA 108, 9513–9515.
Dzialowski, E.M., 2005. Use of operative temperature and standard operative temperature models in thermal biology. J Therm Biol 30, 317–334.
Ghalambor, C.K., Huey, R.B., Martin, P.R., Tewksbury, J.J., Wang, G., 2006. Are mountain passes higher in the tropics? Janzen's hypothesis revisited. Integr Comp Biol 46, 5–17.
Glaw, F., Vences, M., 2007. A field guide to the amphibians and reptiles of Madagascar.
3rd edn, Vences & Glaw, Cologne, Germany.
43 Thermal constraints along an environmental gradient
Goodman, S.M., 1999. A floral and faunal inventory of the Réserve Naturelle Intégrale d'Andohahela, Madagascar: with reference to elevational variation. Fieldiana Zoology New Series 85, 1–319.
Grbac, I., Bauwens, D., 2001. Constraints on temperature regulation in two sympatric Podarcis lizards during autumn. Copeia, 178-186.
Guisan, A., Thuiller, W., 2005. Predicting species distribution: offering more than simple habitat models. Ecol Lett 8, 993–1009.
Gunderson, A.R., Stillman, J.H., 2015. Plasticity in thermal tolerance has limited potential to buffer ectotherms from global warming. P R Soc B 282, 20150401.
Gvozdik, L., 2002. To heat or to save time? Thermoregulation in the lizard Zootoca vivipara (Squamata: Lacertidae) in different thermal environments along an altitudinal gradient. Can J Zool 80, 479–492.
Gvozdik, L., 2012. Plasticity of preferred body temperatures as means of coping with climate change? Biol Lett 8, 262–265.
Gvozdik, L., Castilla, A.M., 2001. A comparative study of preferred body temperatures and critical thermal tolerance limits among populations of Zootoca vivipara (Squamata: Lacertidae) along an altitudinal gradient. J Herpetol 35, 486–492.
Huey, R.B., Deutsch, C.A., Tewksbury, J.J., Vitt, L.J., Hertz, P.E., Perez, H.J.A., Garland, T., 2009. Why tropical forest lizards are vulnerable to climate warming.
P R Soc B 276, 1939–1948.
Huey, R.B., Kearney, M.R., Krockenberger, A., Holtum, J.A.M., Jess, M., Williams, S.E., 2012. Predicting organismal vulnerability to climate warming: roles of behaviour, physiology and adaptation. Philos T R Soc B 367, 1665–1679.
Huey, R.B., Kingsolver, J.G., 1989. Evolution of thermal sensitivity of ectotherm performance. Trends Ecol Evol 4, 131–135.
Huey, R.B., Niewiarowski, P.H., Kaufmann, J., Herron, J.C., 1989. Thermal biology of nocturnal ectotherms - is sprint performance of geckos maximal at low body temperatures. Physiol Zool 62, 488–504.
44
Thermal constraints along an environmental gradient
Janzen, D.H., 1967. Why mountain passes are higher in tropics. Am Nat 101, 233–249.
R Kearney, M., 2013. Activity restriction and the mechanistic basis for extinctions under climate warming. Ecol Lett 16, 1470–1479.
Kearney, M., Porter, W.P., 2004. Mapping the fundamental niche: physiology, climate, and the distribution of a nocturnal lizard. Ecology 85, 3119–3131.
Kearney, M., Shine, R., Porter, W.P., 2009. The potential for behavioral thermoregulation to buffer "cold-blooded" animals against climate warming. P Natl Acad Sci USA 106, 3835–3840.
Leal, M., Gunderson, A.R., 2012. Rapid change in the thermal tolerance of a tropical lizard. Am Nat 180, 815–822.
Levesque, D.L., Nowack J., Stawski C., 2016. Modelling mammalian energetics: the heterothermy problem. Clim Chang Resp 3, 7.
Lighton, J.R.B., 2008. Measuring metabolic rates: a manual for scientists. Oxford University Press, New York, USA.
Little, A.G., Seebacher, F., 2016. Acclimation, acclimatization, and seasonal variation in amphibians and reptiles. In Andrade, D.V., Bevier, C.R. & Carvalho J.E. (Eds).
Amphibian and reptile adaptations to the environment: interplay between physiology and behavior. pp. 41–62. CRC Press, Boca Raton, USA.
Lovegrove B.G., 2009. Modification and miniaturization of Thermochron iButtons for surgical implantation into small animals. J CompPhysiol B 79, 451–458.
Niewiarowski, P.H., Waldschmidt, S.R., 1992. Variation in metabolic rates of a lizard - use of SMR in ecological contexts. Funct Ecol 6, 15–22.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., and R Development Core Team, 2015.
nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-122.
http://CRAN.R-project.org/package=nlme